Self-host sidix
v0.7.0Self-hosted AI agent with self-learning and IHOS
- Stars
- 25
- Requires
- Qwen2.5-7B
What is sidix?
SIDIX is a self-hosted AI agent built on Qwen2.5-7B with LoRA adapters that emphasizes self-learning, provenance tracking, and an epistemological architecture called IHOS (Islamic Holistic Ontological System). Unlike typical chatbots, SIDIX operates as a three-layer cognitive agent: a local LLM brain (Layer 1), 35 tools including web search, code execution, image generation, and multi-agent orchestration (Layer 2), and a continuous learning loop that ingests new corpus data and retrains LoRA adapters quarterly (Layer 3). Key capabilities include the Raudah Protocol for parallel multi-agent research, 5 locked personas (AYMAN, ABOO, OOMAR, ALEY, UTZ) that route based on query type, Sanad chain-of-transmission labeling for every output, Maqashid intent-based filtering, and Naskh knowledge conflict resolution. The project optionally exposes an MCP plugin server for integration with clients like Claude Desktop and Cursor. The hosted service was discontinued in August 2026; users download and run the code themselves. A clean, documented package for self-hosting is being prepared. The research continues in the MiganCore project.
Should you self-host sidix?
sidix scores 31/100 on selfhost.directory (how scoring works) and ranks #7 of 7 in Local AI. It has 25 GitHub stars.
It shipped 1 stable release in the past year. The latest, v0.7.0, came out on 22 April 2026.
Setup is advanced: plan for several services and some ongoing maintenance. It is MIT-licensed and written mostly in Python.
Its closest alternative on selfhost.directory is unsloth (85/100, 78 673 stars). It scores higher overall, so compare the two before you pick. See sidix vs unsloth side by side.
Features
- 35 built-in tools
- 5 locked personas
- Chrome extension (Social Radar)
- Intent-based filtering (Maqashid)
- Local image generation (SDXL)
- Optional MCP plugin serverDisabled by default
- Self-hosted AI Agent
- Self-learning (LoRA retraining)
- Typo-resilient Indonesian NLP
- WhatsApp IntegrationOptional bridge required
sidix README
More in Local AI
SkalesSelf-hosted AI agent harness: background goals, coding, voice and browser automation on desktop plus native phone apps, with 27 providers or fully local inference.
SNDR Core EngineSelf-hosted vLLM patch overlay running Qwen3.6 and Gemma 4 at 280K context on 24 GB GPUs with 1.5× throughput, working tool calls, and neural-graph memory.
Text Generation WebUIAn open-source, self-hosted LLM chat app you run on your own machine: vision input, tool-calling, LoRA fine-tuning, and an OpenAI/Anthropic-compatible API.
Turbo-LLMSelf-hosted local LLM platform running any engine with auto-tuning, real measured tokens/sec, and OpenAI/Anthropic APIs for Claude Code.
unslothSelf-hosted, open-source UI and framework to run, fine-tune and export LLMs, GGUF/MLX and diffusion models on your own GPU or Mac.
vllm-qwenSelf-hosted OpenAI-compatible inference server running Qwen3.6-27B BF16 on AMD Ryzen AI Max+ 395 with vision support, 256K context, and separated reasoning output.
voice-chat-aiSelf-hosted open-source voice chat app for talking to AI characters in a web UI or terminal, mixing Ollama and cloud LLMs with Spark-TTS, Kokoro or ElevenLabs voices.
AgenticSeekSelf-hosted, open-source alternative to Manus AI: a voice-enabled agent that browses the web, writes and runs code, and plans tasks on local LLMs.
Agentic SignalSelf-hosted visual workflow automation with local AI via Ollama. Drag-and-drop nodes for data, APIs, RAG — no cloud dependencies.
AgenvoySelf-hosted AI agent running locally, auto-generates tools, shares via MCP with Claude Code and Codex, keeps data private.
Agora CosmicaSelf-hosted, open-source education platform where 30 historical figures teach through narrated stories, teachings, debates, and live AI conversation.
AI GridAI Grid is a self-hosted, open-source orchestrator that pools your computers into a single OpenAI-compatible endpoint for local LLM inference and training.
0 comments